1. Field of the Invention
The present invention relates to an apparatus for recording and reproducing digital video signals after bit rate reduction.
2. Description of the Prior Art
In a conventional digital audio apparatus or digital video apparatus, copying has been performed by using the audio signals or video signals after converting them to digital signals. What provides the greatest problem in recording and reproducing the digital signals is the fact that the data amount thereof is very large. To meet this problem, various bit rate reduction systems have been proposed, and recording and reproducing apparatuses using bit rate reduction have been in development. As an example of the bit rate reduction, there is a system based on variable length coding of the components obtained by subjecting the objective signal to an orthogonal transformation. In the orthogonal transformation and the inverse orthogonal transformation, errors arise which are caused by rounding in view of the number of digits used for operation thereof and the number output digits. For example, in the 8.times.8 Hadamard transformation, when the accuracy of the input is 8 bits, the accuracy of the output is 14 bits, but what is actually used for bit rate reduction is the upper 9-10 bits. Also, in the 8.times.8 discrete cosine transformation (normally abbreviated as DCT), even when the accuracy of the input is 8 bits and the accuracy of the output coefficient is more than 16 bits, what is used for the bit rate reduction is about the upper 10 bits. Moreover, the real value of the coefficient is an irrational number. In the copying method as above, due to the repetition of the inverse orthogonal transformation of the reproduction of the copy output and the orthogonal transformation of the recording of the copy input on each copying, there is a problem of an accumulation of distortion.
Also, as the video signals are in general transmitted or recorded by dividing data into specified blocks, with respect to the errors which occurred at the channel, the errors could be detected in block units but could not be corrected. In the conventional practice, replacement was made by a block which is 1 page (1 page represents one field or plural fields) before that of the related block for concealment. As the video signal has an extremely large inter-page correlation, it was possible to reproduce efficiently the blocks with which the errors were detected by such concealment. However, in the constitution as above, when the scene changes from that of one page before, due to the decrease of the inter-page correlation, there is a possibility for a large degradation to occur by concealment. Also, in case of an increase in the blocks in which only detection of the errors that occurred as a result of the repetition of transmission or recording was effected, there occurred blocks which were further concealed by using the concealed blocks, thereby incurring a large degradation of the image quality in the rapidly moving video signals.